Food Chemistry 169 (2015) 141–149

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Optimisation and validation of the microwave-assisted extraction of phenolic compounds from rice grains W. Setyaningsih a,b, I.E. Saputro b, M. Palma b,⇑, C.G. Barroso b a b

Department of Food and Agricultural Product Technology, Faculty of Agricultural Technology, Gadjah Mada University, Jalan Flora, 55284 Yogyakarta, Indonesia Department of Analytical Chemistry, Faculty of Sciences, University of Cadiz, Campus de Excelencia Internacional Agroalimentario (CeiA3), Spain

a r t i c l e

i n f o

Article history: Received 10 September 2013 Received in revised form 22 July 2014 Accepted 29 July 2014 Available online 7 August 2014 Keywords: Microwave-assisted extraction Phenolic compounds HPLC-PDA Rice grains

a b s t r a c t A new microwave-assisted extraction (MAE) method has been investigated for the extraction of phenolic compounds from rice grains. The experimental conditions studied included temperature (125–175 °C), microwave power (500–1000 W), time (5–15 min), solvent (10–90% EtOAc in MeOH) and solvent-tosample ratio (10:1 to 20:1). The extraction variables were optimised by the response surface methodology. Extraction temperature and solvent were found to have a highly significant effect on the response value (p < 0.0005) and the extraction time also had a significant effect (p < 0.05). The optimised MAE conditions were as follows: extraction temperature 185 °C, microwave power 1000 W, extraction time 20 min, solvent 100% MeOH, and solvent-to-sample ratio 10:1. The developed method had a high precision (in terms of CV: 5.3% for repeatability and 5.5% for intermediate precision). Finally, the new method was applied to real samples in order to investigate the presence of phenolic compounds in a wide variety of rice grains. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Cultivated rice (Oryza sativa L.) is the second most widely grown cereal crop in the world and it serves as an important staple food for more than half of the world’s population (OECD-FAO. Agricultural Outlook, 2011). Besides the contribution of rice to the total human calorie intake, rice contains some specific components with proven benefits for human health. Several compounds with antioxidant activity have been identified in rice and these include phenolic compounds (Zhou, Robards, Helliwell, & Blanchard, 2003). Phenolic compounds have health benefits due to their antioxidant activities, which have inhibitory effects on mutagenesis and carcinogenesis (Vattem, Ghaedian, & Shetty, 2005). The most common forms of phenolic compounds in rice are represented by hydroxycinnamic and hydroxybenzoic acids. The predominant phenolic compounds in rice are ferulic and p-coumaric acids, both of which are hydroxycinnamic acids (Li, Friel, & Beta, 2010; Lin & Lai, 2011). Other compounds identified include sinapic, p-hydroxybenzoic and protocatechuic acids, which are benzoic acids (Goffman & Bergman, 2004; Vichapong, Sookserm, Srijesdaruk, Swatsitang, & Srijaranai, 2010; Walter & Marchesan, ⇑ Corresponding author. Tel.: +34 956 016 360; fax: +34 956 016 460. E-mail address: [email protected] (M. Palma). http://dx.doi.org/10.1016/j.foodchem.2014.07.128 0308-8146/Ó 2014 Elsevier Ltd. All rights reserved.

2011). In addition to the aforementioned two major groups of compounds, aldehyde analogs such as vanillin are also referred to as phenolics (Qiu, Liu, & Beta, 2010). The concentrations of these compounds in rice grains have been positively correlated with antioxidant activity (Sompong, Siebenhandl-Ehn, Linsberger-Martin, & Berghofer, 2011; Yafang, Gan, & Jinsong, 2011). Thus, phenolic compounds contribute to the antioxidant activity of rice grains. Reversed phase-high performance liquid chromatography (RPHPLC) has been the most commonly used analytical technique for the qualitative and quantitative analysis of phenolic compounds (Duckstein & Stintzing, 2011; Sahin, Demir, & Malyer, 2011). HPLC coupled with a photodiode-array detector (PDA) is sufficiently sensitive for the detection of eluted compounds at wavelengths across the UV and visible spectrum. This technique provides both a sensitive quantitative response and qualitative information regarding the UV–vis chromophore of phenolic compounds. Extraction is a very important analytical step in the isolation and identification of compounds from solid samples prior to chromatographic determination. The development of an optimal procedure for the extraction of phenolic compounds from food samples can prove difficult due to the structural diversity of phenolic compounds and their potent antioxidant activity, which can lead to rapid reaction with other constituents in the matrix (Ajila, Brar, & Verma, 2011). Microwave-Assisted Extraction (MAE) appears to

142

W. Setyaningsih et al. / Food Chemistry 169 (2015) 141–149

be one of the best methods to extract phenolic compounds due to the special microwave/matter interactions and the very rapid extraction time (Li, Skouroumounis, Elsey, & Taylor, 2011; Mandal, Mohan, & Hemalatha, 2007). The MAE system rapidly generates heat and this characteristic results in a shorter extraction time and good quality extracts with better target compound recovery (Barbero, Palma, & Barroso, 2006; Liazid, Guerrero, Cantos, Palma, & Barroso, 2011). Different chemical substances absorb microwaves to different extents and this behaviour makes MAE an efficient method for extractions and, more importantly, it makes it possible to selectively extract target compounds from complex food matrices (Eskilsson & Björklund, 2000; Hemwimon, Pavasant, & Shotipruk, 2007). However, the use of MAE has not previously been reported for the extraction of phenolic compounds from rice grains. The efficiency of the MAE process depends on extraction time, extraction temperature, solid–liquid ratio and the type and composition of solvent used (Pizarro, Pérez-del-Notario, & GonzálezSaiz, 2007; Rostagno, Palma, & Barroso, 2007; Song, Li, Liu, & Zhang, 2011). Chemometric approaches based on the use of an experimental design have been successfully utilised to evaluate the variables that affect MAE recoveries (Zhong et al., 2010). This statistical and mathematical technique has been used to develop, improve and optimise processes (Tabaraki & Nateghi, 2011; Prasad et al., 2011). Previously used extraction methods for phenolics in rice grains usually involve solid–liquid maceration with long extraction times (up to 1 h) (Qiu et al., 2010) and they also require some additional steps, including solvent removal by evaporation to dryness, after completion of the extraction process (Shao, Xu, Bao, & Beta, 2014). The aim of the study reported here was to optimize a rapid MAE method for the extraction of phenolic compounds from rice grains by response surface methodology (RSM).

2. Materials and methods 2.1. Materials and chemicals HPLC-grade methanol (MeOH), acetic acid and ethyl acetate (EtOAc) were purchased from Merck KGaA (Darmstadt, Germany). Phenolic compound standards of the highest available purity were used. Protocatechuic acid (PRO), vanillin (VAN), protocatechuic aldehyde (PARA), p-hydroxybenzoic acid (p-HBA), p-hydroxybenzaldehyde (p-HB), ferulic acid (FER) and sinapic acid (SIN) were obtained from Fluka (Buchs, Switzerland). Guaiacol (GUA), p-Coumaric acid (p-COU), caffeic acid (CAF), 5-hydroxymethyl-2-furaldehyde (HMF), furfural (FUR), 5-methylfurfural (MF) and syringic acid (SYR) were obtained from Sigma–Aldrich (St. Louis, MO, USA). Ellagic acid (ELL) was purchased from Sarsynthese (Merignac, France). Water was purified with a Milli-Q purification system (Billerica, MA, USA).

2.2. Rice samples Rice samples were obtained from a commercial market in Spain. Each rice sample (20 g) was placed in a plastic cylinder and rice grains were milled with an Ultraturrax homogenizer (IKAÒ T25 Digital, Germany) for 10 min prior to extraction. The milling process was stopped every 1 min to avoid excessive heating of the sample. The fine grain was then homogenised by stirring and then stored in a closed bottle. The final extraction method was applied to 13 different rice products available in the market and these covered the varieties of short (4), long (3), aromatic (2), exotic (1), whole grain (2) and processed (1) rice grains.

2.3. Extraction of phenolic compounds MAE experiments were performed with a Milestone Ethos 1600 system (Sorisole, Italy) equipped with vessels made from tetrafluoromethoxyl polymer with Teflon liners. Rice powder (2.5 g) was accurately weighed and placed into an extraction vessel. According to the experimental design, a set volume and type of solvent was added to the extraction vessel and the extraction was performed under different MAE conditions. After extraction, the vessels were cooled in an ice bath for 10 min and carefully opened in a fume cupboard. The solid material in the sample was filtered off and washed using fresh solvent. The combined filtrate and washings were evaporated to dryness under vacuum (rotary evaporator). The residue was re-dissolved in methanol (2 mL) and was filtered through a 0.45 lm filter prior to injection into the HPLC-FD system. 2.4. Determination of phenolic compounds The HPLC system comprised a Dionex P680 HPLC Pump, Dionex ASI 100 Automated Sample Injector, Dionex PDA-100 Photodiode Array Detector, Dionex UCI-50 Universal Chromatography Interface, and Dionex TCC-100 Thermostatted Column Compartment. Separations were performed on a reversed phase RP 18 LiChrospher Column (LiChroCART 250  4 mm (5 lm), Merck KgaA, Darmstadt, Germany). Gradient elution was carried out at a flow rate of 1.0 mL min1. A PDA-100 Photodiode Array Detector was used for UV–vis measurements and the 3D mode was set at collection rate of 1.0 Hz, 3D wavelength scan range of 250–600 nm, 3D bunch width of 1 nm and band width of 50 nm. The column compartment thermostat was set at 25 °C. The injection volume was set to 25 lL. A gradient elution was programmed using an acidified aqueous mobile phase A (2% acetic acid and 5% methanol) and mobile phase B (2% acetic acid and 88% methanol). The gradient applied was as follows: (time, solvent B): 0 min, 0%; 10 min, 25%; 25 min, 40%; 30 min, 50%; 35 min, 50%. The identification of phenolic acids in the samples was achieved by spiking and by comparison of retention times and maximum UV absorptions with those of standards. The analytical properties of the chromatographic method for the determination of 15 phenolic compounds are listed in Table 1. Typical chromatograms at three different wavelengths (260, 280 and 320 nm) are shown in Fig. 1. 2.5. Performance of the method The chromatographic analytical procedure used to determine phenolic compounds was carried out according to the ICH Guideline Q2 (R1) and suggestions made in ISO 17025 (ICH, 2006; ISO, 2005). The linearity, range, precision, detection and quantification limits of the method were evaluated. Linearity was assessed in order to confirm the ability of the method to obtain test results that are directly proportional to the concentration of phenolic compounds within the range studied. Stock solutions of each phenolic compound were diluted as appropriate to give concentrations ranging from 0.15 to 30 mg L1. Gnumeric 1.10.17 was used to generate the regression analysis to obtain the calibration curves and quantify the phenolic compounds in the extracts. The standard deviation (r) obtained for the response and the slope (m) from the regression were then used to calculate the limit of detection (LOD) and limit of quantification (LOQ) using Eq. (1) and Eq. (2), respectively.

LOD ¼ 3:3 r=m

ð1Þ

LOD ¼ 10 r=m

ð2Þ

143

Peak area

3.3 0.76 0.36 1.9 1.04 0.36 1.90 1.18 0.86 2.09 1.28 0.99 1.27 1.34 2.17

Peak height

0.6 2.23 0.45 1.85 2.22 0.63 1.85 1.36 0.56 1.83 0.71 2.19 2.10 1.65 3.43

tR

0.21 0.26 0.44 0.18 0.21 0.06 0.18 0.48 0.17 0.22 0.24 0.16 0.14 0.14 0.31 0.37 1.39 0.07 1.28 1.71 0.23 1.97 0.27 0.12 2.04 2.57 1.87 1.92 1.72 1.15 0.17 0.33 0.17 0.33 0.38 0.4 0.49 0.42 0.34 0.48 1.43 0.36 0.65 0.45 3.26 0.16 0.00 0.23 0.00 0.00 0.05 0.00 0.06 0.14 0.00 0.00 0.04 0.04 0.04 0.05 0.18 0.23 1.24 0.16 0.22 0.83 0.22 0.93 0.16 0.20 0.37 0.18 0.27 0.26 1.59 0.9998 0.9997 0.9919 0.9999 0.9997 0.9963 0.9997 0.9954 0.9998 0.9998 0.9993 0.9998 0.9996 0.9996 0.9936

0.61 0.76 4.14 0.55 0.74 2.79 0.74 3.11 0.55 0.66 1.24 0.61 0.90 0.89 4.82

Peak area Peak height tR

LOD (mg L1) R2

The precision of the method was evaluated by studying repeatability (intra-day) and intermediate precision (extra-day). Repeatability was assessed by nine independent injections of samples on the same day while intermediate precision was determined by three independent HPLC analyses on three consecutive days. Precision was expressed as Coefficient of Variance (CV) of the retention time and peak height. The acceptable CV limit was ±10% according to the AOAC manual for the Peer-Verified Methods program (AOAC, 1993). The CV values with reference to the height of the peak for both repeatability (0.8%) and intermediate precision (1.5%) were less than 2%, thus showing that the method has excellent precision. 2.6. Experimental design The response surface methodology was employed to explore the variables that affect the microwave extraction and this approach enables the overall number of experiments and possible interactions between the variables to be considered. A central composite design was developed and a fractional three-level/five-factor experimental design with three replicates at the centre point was used to investigate the effects of five independent variables on the extraction of phenolic compounds from rice samples. The independent variables were coded at three levels (1, 0 and 1) and each level was selected on the basis of the stability of the phenolic compounds under MAE conditions, as described by Liazid, Palma, Brigui, and Barroso (2007). The chosen levels for the variables and the fractional factorial design are provided in Table 2. The complete experimental design consisted of 29 experimental points (Table 2). The technique described above was used to obtain the surface response by fitting the data to a polynomial model and the effects of each factor and also the interactions between factors were evaluated. The most general function for the central composite design is represented in Eq. (3).

Linear equation

Y = 1.32X  1.95 Y = 1.32X  0.05 Y = 1.32X + 2.37 Y = 5.52X + 0.03 Y = 7.43X  0.23 Y = 1.32X  1.05 Y = 3.44X  0.07 Y = 1.32X + 0.28 Y = 1.32X + 14 Y = 1.85X  0.02 Y = 0.35X + 0.03 Y = 7.52X + 0.16 Y = 4.96X  0.27 Y = 1.32X  0.29 Y = 0.03X + 0.02

Range (mg L1)

0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12 0.15–12

Y ¼ b0 þ

Xn

bX i¼1 i i

þ

Xn Xn i¼1

bXX j¼1 i i j

ð3Þ

where Xi are the studied variables (temperature (°C), X1; microwave power (Watts), X2; solvent (percentage of EtOAc in mixtures with MeOH), X3; sample:solvent ratio (g of sample/mL of solvent), X4; and extraction time, X5 (min)); the response Y was the amount of phenolic compounds obtained in the extraction. The b coefficients are obtained by the partial least square method. 2.7. Data analysis The experimental results from single factor experiments were analysed using Gnumeric. The Analysis of Variance (ANOVA) and Least Significant Difference (LSD) test were used to determine the significant differences between the means. The construction and analysis of the experimental design, the response surface, and the desirability functions to achieve the optimum conditions were obtained using Unscrambler X (CAMO) (Oslo, Norway). 3. Results and discussion

5-HMF (1) Protocatechuic acid (2) Furfural (3) Protocatechuicaldehyde (4) p-OHBenzoic acid (5) Caffeic acid (6) p-OHBenzaldehyde (7) 5-Methylfurfural (8) Syringic acid (9) Vanillin (10) Guaiacol (11) p-Coumaric acid (12) Ferulic acid (13) Sinapic acid (14) Ellagic acid (15)

3.1. Effects of the extraction variables

Phenolic compounds

Table 1 Analytical characteristics for determination of phenolic compounds.

LOQ (mg L1)

Intra-day precision, CV (%) n = 9

Inter-day precision, CV (%) n = 3  3

W. Setyaningsih et al. / Food Chemistry 169 (2015) 141–149

An investigation into the stability of phenolic compounds under MAE conditions was carried out by Liazid et al. (2007) and this proved useful to identify the appropriate variables and working range to optimise the MAE. In this respect, the variables likely to influence the extraction include temperature (X1; 125–175 °C), microwave power (X2; 500–1000 W), solvent (X3; 10–90% EtOAc), solvent-to-sample ratio (X4; 20:1–10:1) and extraction time (X5; 5–15 min). Since the variables have different units and ranges,

144

W. Setyaningsih et al. / Food Chemistry 169 (2015) 141–149

Fig. 1. HPLC-PDA chromatograms of phenolic compounds in a rice extract sample. See Table 1 for peak identities.

each of the variables must first be normalised and forced into a range from 1 to + 1 in order to obtain a more even response (Bas & Boyacı, 2007).

The experimental design was developed using Unscrambler X software to determine the effect of process variables, each at three equidistant levels (1, 0,+1), and the interactions between these

Table 2 Central composite design of five variables with their observed responses. Exp. no

Extraction variables X1

X2

X3

X4

X5

Temperature (°C)

Power (W)

Time (min)

Solvent (% EtOAc)

Solvent to sample ratio

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

1 1 1 1 0 1 1 1 1 1 1 1 1 0 1 0 0 1 1 1 1 1 1 1 1 1 0 0 0

1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 0 0 0

1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 0 1 1 1 1 1 1 1 0 0 0 0

1 1 1 1 0 1 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0

1 1 0 1 1 1 1 1 1 0 1 1 0 1 1 1 1 1 1 1 0 1 1 1 0 1 0 0 0

125 175 175 125 150 175 125 125 175 175 125 175 125 150 125 150 150 175 175 175 125 125 175 175 125 125 150 150 150

500 500 1000 1000 1000 1000 1000 500 1000 500 750 1000 1000 1000 500 1000 500 500 1000 750 500 500 500 500 1000 750 750 750 750

5 5 15 5 15 5 5 15 15 15 15 15 10 10 15 5 15 10 5 5 5 5 5 15 15 10 10 10 10

10 10 90 10 50 10 50 50 50 90 90 10 90 10 10 90 90 90 90 90 90 10 90 10 10 90 50 50 50

20 10 15 20 20 10 10 10 10 15 10 20 15 10 20 10 20 10 20 10 15 10 20 10 15 20 15 15 15

Total phenolic (lg/g)

Predicted relative values

Relative error of prediction (%)

6.49 5.29 4.27 1.75 4.24 6.13 1.94 3.10 9.73 4.13 1.39 10.75 0.72 4.63 2.60 1.45 2.01 2.97 2.30 2.21 0.39 2.34 1.12 10.27 2.88 0.45 3.29 3.77 3.28

2.68 6.39 4.58 2.74 4.64 6.75 1.48 3.13 7.19 4.45 1.18 8.87 0.13 5.92 4.50 1.17 2.20 3.81 2.21 2.97 0.64 2.94 2.35 8.54 4.99 0.47 3.81 3.81 3.81

83 19 7 44 9 10 27 1 30 7 16 19 138 24 54 21 9 25 4 29 48 23 71 18 54 4 15 1 15

W. Setyaningsih et al. / Food Chemistry 169 (2015) 141–149 Table 3 Regression coefficients for response in the regression analysis. Model term

Estimate

p-Value

b0 b1 b2 b3 b4 b5 b12 b13 b23 b14 b24 b34 b15 b25 b35 b45 b11 b22 b33 b44 b55

0.2244 0.1296 0.0232 0.0884 0.1371 0.0334 0.0827 0.1335 0.0949 0.1107 0.1132 0.0667 0.0231 0.0001 0.0495 0.1332 0.1304 0.0276 0.0937 0.1388 0.0294

0.00001 0.58286 0.08704 0.00012 0.53396 0.01842 0.00080 0.19635 0.00644 0.00348 0.16751 0.24389 0.30381 0.99241 0.00262 0.00001 0.69801 0.04200 0.00002 0.67345

variables and the response variables. The response for each extraction in the experimental design was calculated as the total amount of phenolic compounds with levels above the limit of quantification (Table 2). The regression coefficient for each of the term combinations of the independent variables was then calculated and the significance was determined using the p-value generated by a ttest. The predicted values for the total amount of phenolics obtained from the model generated using the significant variables (Table 3) are shown in Table 2. The significance of the effects of the independent variables and their interactions on the dependent variable was checked by analysis of variance (ANOVA) and the results are listed in Table 3. The variables that gave rise to the main effects were temperature (X1), time (X3) and solvent (X4) and these are highly significant (p < 0.1). In contrast, the microwave power (X2) and solvent-to-sample ratio (X5) did not have a significant effect (p > 0.1).

145

The interaction between the extraction variables temperature and microwave power (X1X2), temperature and time (X1X3), temperature and solvent (X1X4), microwave power and solvent (X2X4) and solvent and solvent-to-sample ratio (X4X5) were found to be highly significant (p < 0.05). As far as quadratic effects of the extraction variables are concerned, phenolic compounds in the extract were significantly affected (p < 0.05) by quadratic terms of extraction temperature (X1X1), extraction time (X3X3) and solvent (X4X4). The other interactions and quadratic terms of extraction variables were not significant (p > 0.1). A positive value for the estimated effect indicates an increase in the extraction yield if the variable is increased to its highest level. A negative value indicates that a better extraction yield is obtained for low values of the variables in question. The positive value for the effect of extraction temperature and time indicate that, within the studied range, the extraction efficiency increases on increasing these variables. The variable solvent showed a significant negative effect, i.e., a lower percentage of EtOAc in MeOH led to a higher yield of phenolic compounds. The extraction variables temperature, time and solvent were identified as the most significant variables that influence the yield of phenolic compounds. As a consequence, these variables would require a final optimisation. 3.2. Predicted second-order polynomial model The values obtained with the prediction model were 1.272 as root mean square error (RMSE) with an R-Squared value of 0.795 for the predicted vs reference values for recoveries. The predictive capability of the regression model was evaluated by considering a plot of the measured vs predicted values (Fig. 2). The high slope values for this plot indicate that the model equations are appropriate to describe the experimental design. The predicted and real values for phenolic recoveries are shown in Table 2. The average error for the prediction was 28%, although only seven experiments gave very high error values (>40%) and most of these were for extraction conditions that led to low recoveries, i.e. values below 3 lg of phenolics per g of sample. Prediction errors for extractions with higher recoveries were much lower, with an average error of only 14%. Therefore,

Fig. 2. The prediction capability of the regression model.

W. Setyaningsih et al. / Food Chemistry 169 (2015) 141–149

90 93.07 ± 4.51ab

Values followed by the same letter are indicated as not significant different (a = 0.05).

ND ND ND TR ND ND ND ND ND TR TR TR TR TR ND ND ND ND ND ND ND ND TR 3.77 ND ND 1.65 1.33 1.13 1.33 1.26 1.30 0.93 1.27 1.27 3.14 3.21 1.47 3.81 TR TR TR TR TR TR ND TR TR 1.82 3.77 TR 1.83 3.90 3.10 2.65 4.80 3.06 2.30 TR 3.32 1.68 15.99 25.38 2.76 13.57 1.26 1.11 1.16 1.16 0.93 1.08 0.77 1.11 0.99 3.48 3.48 1.19 3.81 TR TR TR 0.60 TR TR TR TR TR TR – TR TR 3.63 7.26 TR TR 4.82 TR TR TR 10.90 – – 4.03 – 2.78 5.52 4.52 3.57 5.01 2.90 4.87 3.69 4.92 5.48 12.97 4.19 5.63 ND ND ND ND ND ND ND ND ND ND ND TR ND TR TR TR TR TR TR TR TR TR TR 4.63 TR ND TR TR ND ND ND ND ND TR ND TR 3.66 ND ND TR ND ND ND TR ND ND ND ND TR ND ND 2.08

VAN SYR MF p-HB CAF p-HBA PRA FUR PRO

Phenolic Compounds (lg/g)**

ND, not detected due to less than LOD. TR, trace due to the concentration less than LOQ but higher than LOD. * Commercial brands of the sample were coded for privacy purpose. ** See experimental section for codes used for phenolic compounds.

30 98.35 ± 3.96bc

Short grain brand A1 Long grain brand B1 Long grain brand B2 Short grain brand A2 Short grain brand A3 Long grain brand B3 Thai Long grain aromatic Basmati long grain aromatic Parboiled long grain Whole short grain brown brand C1 Exotic rice grain Short grain brand A4 Whole short grain brown brand C2

a,b,c

95 91.55 ± 2.12b

25 96.17 ± 2.80bc

1 2 3 4 5 6 7 8 9 10 11 12 13

Solvents (% methanol) 100 100 ± 1.64a

20 100 ± 1.79c

Variety of rice grain*

Extraction time (min) 10 15 82.89 ± 13.01a 89.35 ± 5.76ab

Sample

Table 4 LSD of the yield for MAE at different extraction time and solvents (mean ± RSD for the relative values to the maximum response (%)).

Table 5 Concentration of phenolic compounds in different types of rice grain samples.

3.3.2. Extraction time The extraction time has a significant positive effect on the extraction yield. Thus, the optimisation process should include longer extraction times than the range considered in the preliminary study. The extraction yields obtained at different times are shown in Table 4. Two-factor ANOVA was used to evaluate the significance of extraction time and type of phenolic compound by considering the yields of individual phenolic compounds. The extraction time was found to have a significant effect on the extraction of phenolics as the Fcalculated value for extraction time (4.46) was higher than

GUA

3.3.1. Extraction temperature The extraction temperature had a significant positive effect on the extraction process and higher temperatures were therefore assessed. The limit of the optimisation was dictated by analyte stability and extraction temperatures of 175–185 °C were employed on the basis of results reported by Liazid et al. (2007) for the stability of phenolic compounds under MAE conditions. The regression equation could not be applied because the best extraction temperature was clearly outside the studied range. As one would expect, increases in temperature led to significant (p < 0.05) increases in response. Thus, 185 °C was set as the optimum value for the extraction temperature. Furthermore, two-factor ANOVA was applied to investigate the effect of temperature on the individual phenolic compounds. The data were grouped by extraction temperature and type of phenolic compound; the dependent variable was the peak height of the signal. The Fcalculated values for factors A and B were higher than the corresponding Fcritical value, which indicates that extraction temperature has a significant effect along with the type of phenolic compounds. This situation occurs because the phenolic compounds with different chemical structures behave differently in terms of stability at higher extraction temperatures. According to Liazid et al. (2011) there is a clear relationship between the chemical structure and the stability of phenolic compounds when studied under different MAE conditions. Moreover, it has been shown that the compounds that contain a greater number of hydroxyl-type substituents are more easily degraded under these temperature conditions.

1.03 1.21 0.84 0.89 1.37 TR TR TR 1.21 1.06 3.77 1.16 1.34

p-COU

FER

SIN

ELL

The coordinates of the stationary point were determined to be outside the design region and, as a consequence, the highest yield of phenolic compounds achieved at the highest temperature and time within the investigated range was not the optimum value for the extraction of phenolic compounds by MAE. For this reason, a further study was required to maximise the recovery and this involved an assessment of the significant variables at levels outside the investigated range.

0.70 0.91 TR 0.95 1.17 TR TR TR 1.08 1.71 3.77 0.90 2.33

3.3. Optimisation of the experimental conditions

14.95 20.44 10.3 13.3 17.62 7.58 6.57 9.39 22.05 32.68 68.41 15.7 34.4

Sum (lg/g)

the model can be used to estimate the response for the purposes of optimisation.

HMF

146

W. Setyaningsih et al. / Food Chemistry 169 (2015) 141–149

Fcritical (2.48). Therefore, LSD was performed to investigate the least significant difference between the yields of phenolic compounds extracted for different extraction times (Table 4). The yield obtained in the MAE for 20 min was higher and was significantly different to that obtained at lower extraction times. However, the results were not significantly different on applying an even longer extraction time. The MAE system allows a total of 10 samples to be extracted simultaneously and the final extraction time per sample would be only 2 min.

3.3.3. Solvent The different aqueous MeOH mixtures have different dielectric properties in MAE (Venkatesh & Raghavan, 2004) and this has a profound effect on the extraction of phenolic compounds (Singh et al., 2011). The effect that the type of solvent has on extraction efficiency in MAE was determined and the extract yields were compared (see Table 4). The highest extraction yield was obtained on using 100% methanol. Two-factor ANOVA was used to study the significance of solvent on the yield of individual phenolic compounds. It was found that changes in the solvent have a significant effect on the extraction of phenolics as Fcalculated (5.25) was higher than Fcritical (5.14). Therefore, LSD was carried out to investigate the significance within the different solvents (Table 4). The selected solvent was 100% MeOH as this provided highest recovery and ease of preparation.

147

3.4. Analytical characteristics of the method The final extraction conditions were as follows: extraction temperature 185 °C, microwave power 1000 W, extraction time 20 min, extraction solvent 100% MeOH, and a solvent-to-sample ratio of 10:1. The precision was calculated for the extraction carried out under these conditions. The precision of the method was evaluated by considering repeatability and intermediate precision, which is expressed as the coefficient of variance (CV). The repeatability was measured on the basis of ten extractions in a single day. The effects of random events on the analytical procedure were evaluated through the intermediate precision in a similar way to repeatability, but in this case the analysis was conducted on data collected from extractions on three different days. The CV for repeatability of the extraction using the developed method was 5.35% while the intermediate precision was 5.47%. It was also observed that PRO has the highest repeatability (CV = 0.13%) and FER has the lowest (CV = 8.08%). These CV values for both repeatability and intermediate precision are within the acceptable limits defined by the AOAC, i.e., ±10% according to the AOAC manual for the Peer-Verified Methods program (AOAC, 1993) and these results show the good precision of the extraction method. 3.5. Application to a real sample The applicability of this MAE technique for rice grains as the matrix samples was evaluated by extracting several samples of rice

Fig. 3. Relative values to the maximum response (%) of classified rice grain samples.

148

W. Setyaningsih et al. / Food Chemistry 169 (2015) 141–149

grains available in Spanish markets using the developed method. The extracts were analysed by HPLC-PDA. The results for the phenolic compounds obtained from these rice samples are shown in Table 5. It was found that exotic rice grains have the highest level (68.41 lg g1) of extracted phenolic compounds whereas aromatic rice grains on average have the lowest (6.57–9.39 lg g1). The amounts of phenolic compounds extracted from regular rice were in the range 13.30–17.62 lg g1 whereas for whole short grain rice the level reached 34.40 lg g1. Rice grain samples were classified in order to simplify the analysis on the basis of the variety and type of processing (see Fig. 3). The highest level of phenolic compounds was obtained from exotic rice, which has black grains from wild rice (Zizania aquatica), in 1:3 mixtures with parboiled rice. Other researchers have also reported that grains with a darker colour contain the highest levels of polyphenols (Tian, Nakamura, & Kayahara, 2004). The phenolic concentration in rice appears to be strongly related to rice colour, with black and brown cultivars showing higher concentrations than those grains that are white or pale-yellow. The second highest levels of phenolic compounds after exotic rice were found in whole grain brown rice. This type of rice was not polished and thus the pericarp was still present. Polishing significantly reduces the concentration of phenolics since these compounds are located mainly in the external layers of the grain (Vichapong et al., 2010; Walter & Marchesan, 2011). Therefore, as the phenolic compounds are mainly associated with the pericarp in rice, the milling process to produce white grain (long and short grain) reduces the concentration of these compounds in the grain. The variation in grain-shape (short and long grain) was not correlated with phenol concentration. This finding is consistent with the results reported by Hak-Soo and Woon-Goo (2011), who concluded that the long-grain and short-grain types did not differ significantly in psycho-chemical properties but only in different lengths and spikelet numbers per panicle. 5-Hydroxymethyl-2-furaldehyde, protocatechuic acid, phydroxybenzaldehyde, vanillin, guaiacol and ferulic acid were detected and quantified in all analysed samples. The phenolics that had the highest levels for all analysed samples were para-hydroxybenzaldehyde and guaiacol. 4. Conclusions The study reported here is, to our knowledge, the first time that MAE has been employed and optimised for the extraction of phenolic compounds from rice grains. The proposed method showed satisfactory linearity (correlation coefficients between 0.997 and 0.999) and precision (in terms of CV, 5.35% for repeatability and 5.47% for intermediate precision). The developed MAE method was applied to real samples of rice grains and satisfactory results were obtained. A reliable chromatographic method was used to show that 15 phenolic compounds (5-hydroxymethyl-2-furaldehyde, protocatechuic acid, furfural, protocatechuicaldehyde, p-hydroxybenzoic acid, caffeic acid, p-hydroxybenzaldehyde, 5-methylfurfural, syringic acid, vanillin, guaiacol, p-coumaric acid, ferulic acid, sinapic acid, and ellagic acid) were present in the extracts of rice grains. It can be concluded from the results that MAE under optimum conditions can be considered as a powerful tool for the determination of phenolic compounds from a wide variety of rice grains. Acknowledgements This Research forms part of the Research Thesis carried out by W.S. at the Erasmus Mundus Master Quality in Analytical Laboratories. W.S. thanks the Education, Audiovisual and Culture Execu-

tive Agency (EACEA) for a fellowship. W.S. also thanks the CIMB Foundation for a Ph.D. studentship through the CIMB Regional Scholarships 2012. References Ajila, C. M., Brar, S. K., & Verma, M. (2011). Extraction and analysis of polyphenols: Recent trends. Critical Review in Biotechnology, 31, 227–249. AOAC (1993). AOAC peer verified methods program, manual on policies and procedures. Arlington, VA: AOAC International. Barbero, G. F., Palma, M., & Barroso, C. G. (2006). Determination of capsaicinoids in peppers by microwave-assisted extraction–high-performance liquid chromatography with fluorescence detection. Analytica Chimica Acta, 578, 227–233. Bas, D., & Boyacı, I. H. (2007). Modeling and optimization I: Usability of response surface methodology. Journal of Food Engineering, 78, 836–845. Duckstein, S. M., & Stintzing, F. C. (2011). Investigation on the phenolic constituents in Hamamelis virginiana leaves by HPLC-DAD and LC–MS/MS. Analytical and Bioanalytical Chemistry, 401, 677–688. Eskilsson, C. S., & Björklund, E. (2000). Analytical-scale microwave-assisted extraction. Journal of Chromatography A, 902, 227–250. Goffman, F. D., & Bergman, C. J. (2004). Rice kernel phenolic content and its relationship with anti-radical efficiency. Journal of Science and Food Agriculture, 84, 1235–1240. Hak-Soo, S. U. H., & Woon-Goo, H. A. (2011). Character variations of Korean weedy rice. Rice Genetic Newsletter, 11, 69. Hemwimon, S., Pavasant, P., & Shotipruk, A. (2007). Microwave-assisted extraction of antioxidative anthraquinones from roots of Morinda citrifolia. Separation and Purification Technology, 54, 44–50. ICH (2006). ICH topic Q 2 (R1) validation of analytical procedures: Text and methodology. London: EMEA. ISO (2005). ISO/IEC 17025:2005-general requirements for the competence of testing and calibration laboratories (2nd ed.). Switzerland: ISO. Li, W., Friel, J., & Beta, T. (2010). An evaluation of the antioxidant properties and aroma quality of infant cereals. Food Chemistry, 121, 1095–1102. Li, Y., Skouroumounis, G. K., Elsey, G. M., & Taylor, D. K. (2011). Microwaveassistance provides very rapid and efficient extraction of grape seed polyphenols. Food Chemistry, 129, 570–576. Liazid, A., Guerrero, R. F., Cantos, E., Palma, M., & Barroso, C. G. (2011). Microwave assisted extraction of anthocyanins from grape skins. Food Chemistry, 124, 1238–1243. Liazid, A., Palma, M., Brigui, J., & Barroso, C. G. (2007). Investigation on phenolic compounds stability during microwave-assisted extraction. Journal of Chromatography A, 1140, 29–34. Lin, P., & Lai, H. (2011). Bioactive compounds in rice during grain development. Food Chemistry, 127, 86–93. Mandal, V., Mohan, Y., & Hemalatha, S. (2007). Microwave assisted extraction – An innovative and promising extraction tool for medicinal plant research. Pharmacognosy Reviews, 1, 7–18. OECD-FAO. Agricultural Outlook 2011, release 6, 2011, . Pizarro, C., Pérez-del-Notario, N., & González-Saiz, J. M. (2007). Optimization of a microwave-assisted extraction method for the simultaneous determination of haloanisoles and halophenols in cork stoppers. Journal of Chromatography A, 1149, 138–144. Prasad, K. N., Hassan, F. A., Yang, B., Kong, K. W., Ramanan, R. N., Azlan, A., et al. (2011). Response surface optimization for the extraction of phenolic compounds and antioxidant capacities of under utilised Mangifera pajang Kosterm. peels. Food Chemistry, 128, 1121–1127. Qiu, Y., Liu, Q., & Beta, T. (2010). Antioxidant properties of commercial wild rice and analysis of soluble and insoluble phenolic acids. Food Chemistry, 121, 140–147. Rostagno, M. A., Palma, M., & Barroso, C. G. (2007). Microwave assisted extraction of isoflavones. Analytica Chimica Acta, 588, 274–282. Sahin, S., Demir, C., & Malyer, H. (2011). Determination of phenolic compounds in Prunella L. by liquid chromatography–diode array detection. Journal of Pharmaceutical and Biomedical Analysis, 55, 1227–1230. Shao, Y., Xu, F., Bao, J., & Beta, T. (2014). Phenolic acids, anthocyanins, and antioxidant capacity in rice (Oryza sativa L.) grains at four stages of development after flowering. Food Chemistry, 143, 90–96. Singh, A., Sabally, K., Kubow, S., Donnelly, D. J., Gariepy, Y., Orsat, V., et al. (2011). Microwave-assisted extraction of phenolic antioxidants from potato peels. Molecules, 16, 2218–2232. Sompong, R., Siebenhandl-Ehn, S., Linsberger-Martin, G., & Berghofer, E. (2011). Physicochemical and antioxidative properties of red and black rice varieties from Thailand, China and Sri Lanka. Food Chemistry, 124, 132–140. Song, J., Li, D., Liu, C., & Zhang, Y. (2011). Optimized microwave-assisted extraction of total phenolics (TP) from Ipomoea batatas leaves and its antioxidant activity. Innovative Food Science and Emerging Technologies, 12, 282–287. Tabaraki, R., & Nateghi, A. (2011). Optimization of ultrasonic-assisted extraction of natural antioxidants from rice bran using response surface methodology. Ultrasonic Sonochemistry, 18, 1279–1286. Tian, S., Nakamura, K., & Kayahara, H. (2004). Analysis of phenolic compounds in white rice, brown rice, and germinated brown rice. Journal of Agricultural and Food Chemistry, 52, 4808–4813.

W. Setyaningsih et al. / Food Chemistry 169 (2015) 141–149 Vattem, D. A., Ghaedian, R., & Shetty, K. (2005). Enhancing health benefits of berries through phenolic antioxidant enrichment: Focus on cranberry. Asia Pacific Journal of Clinical Nutrition, 14, 120–130. Venkatesh, M. S., & Raghavan, G. S. V. (2004). An overview of microwave processing and dielectric properties of agrifood materials. Biosystems Engineering, 88, 1–18. Vichapong, J., Sookserm, M., Srijesdaruk, V., Swatsitang, P., & Srijaranai, S. (2010). High performance liquid chromatographic analysis of phenolic compounds and their antioxidant activities in rice varieties. Food Science and Technology, 43, 1325–1330. Walter, M., & Marchesan, E. (2011). Phenolic compounds and antioxidant activity of rice. Brazilian Archives of Biology and Technology, 54, 371–377.

149

Yafang, S., Gan, Z., & Jinsong, B. (2011). Total phenolic content and antioxidant capacity of rice grains with extremely small size. African Journal of Agricultural Research, 6, 2289–2293. Zhong, M., Huang, K., Zeng, J., Li, S., She, J., Li, G., et al. (2010). Optimization of microwave-assisted extraction of protopine and allocryptopine from stems of Macleaya cordata (Willd) R. Br. using response surface methodology. Journal of Separation Science, 33, 2160–2167. Zhou, Z., Robards, K., Helliwell, S., & Blanchard, C. (2003). The distribution of phenolic acids in rice. Food Chemistry, 87, 401–406.